i           home >
s e l e c t Masako Sho | Chaitee Sengupta
 
Project Prospectus
  1. Problem | 2. Analysis | 3. Suggested Solution
4. Plan | 5. Participating Students
  1. Problem
  Images play a very important role in various courses on campus, such as Art, History, Journalism and Architecture. Traditionally, images are stored in the form of slides which are housed in various campus slide libraries. Access to these slides is quite limited. Generally, only faculty have privileges to check them out. Undergraduate students rarely get access to them outside of lecture, when they are presented in slide shows.
   
  In recent years there have been several efforts to digitize instructional image collections and use the digitized images. The digitized images are stored in various databases. Some of these database are searchable over the web. However generally these sites have not been used directly in instruction.
   
  At the same time, there have been efforts on campus to use web technology to augment instruction. Many professors in image-intensive disciplines have taken the initiative to create web sites for their courses. They have done this through various means, such as creating them themselves, or with the assistance of support people in their departments, working with the Instructional Technology Program/Faculty Internet Service Center, the Office of Media Services, the Berkeley Multimedia Research Center or the SIMS Course Website Design project. Web sites have been created for individual courses on which around 200 images have been displayed, along with some textual information.
   
  These web sites are especially useful for students who used to have only limited access to these images through the library in the form of books and slides. However, these images on the course web sites are still similar to slides and books in the sense that viewers can do nothing but browse and print.
   
  The static aspect of course web sites with images also creates difficulty on the publishing side. That is, it is troublesome to insert or delete images from existing set of images or change their order on a page because such modification requires tedious revision of HTML code. Furthermore, creating static web sites with hundreds of image files uses quite a bit of disk space. It is also very difficult to re-use these images because they are "trapped" within whatever site they belong to.
   
  Furthermore, while many images are available online, instructors of such courses still select images for these web sites, lecture, paper or course assignments from slides stored in boxes or books. In spite of the availability of digital images, interfaces for image database are not optimized for the use by instructors and students in the same or more efficient way than with analog media.
  ^ top
  2. Analysis
  Three groups on campus, faculty, students and webmasters, have different kinds of problems in the process starting from image selection to the use of course web sites.
   
 

Faculty

  • The way to make queries to an image database using metadata does not map well to the way instructors select images from slides or books
  • It is difficult or impractical to see an overview of available images before making queries.
  • They cannot select and keep a set of images for later use, such as lecture and course web sites development.
   
 

Students

  • If images related to their coursework are not available on the web, students have only limited access to valuable resources through analog media.
  • Even when images are available on the web, students can view them only in the way they have been already categorized and arranged. For example, it is difficult to compare two images on the screen when they are placed in separate web pages.
   
 

Webmasters

  • Without domain knowledge of particular images, it is difficult for webmasters to name, categorize and arrange images according to instructors' intention. Faculty is generally too busy to spend much time with web designers to share important information. Designers must often make guesses about subject matter when designing, risking faculty disappointment.
  • Adding new images to a course site and changing the order of images on existing pages can be very tedious and time consuming tasks if the site is made of static HTML pages.
  • It is very tedious to create an image-intensive site. Images must be scanned, manipulated (create various sizes), named logically, and associated with accurate textual information.
  ^ top
  3. Suggested Solution
  We are interested in providing a solution that covers the entire process starting from the image selection to the use of course web sites with images. The current process is segmented into four parts, image selection and image grouping, site development, and image access involving faculty, webmasters and students respectively.
 
current
suggested
  Our solution will integrate image selection, image grouping, and site development parts and add personalization to the image access phase of the process. It is intended for:
  image selection: Our system lets instructors select images from databases by both browsing and searching and keep a set of images for later use.
  HTML generation: Course web site with selected images are dynamically generated.
  image access: Selected images are available to students via the Internet.
  personalization: Students can personalize the course web site by selecting particular images, changing the order of images, or adding texts to them.
   
  Our system will provide solutions to the problems presented in the previous section.
 

Faculty

  • The browsing function that enables users to "pick up" images while viewing a large number of images provides users with an experience which more closely maps to selecting slides in boxes or choosing images from books.
  • Browse and search functions provide ways to change the direction of a search at any time. Since both functions work seamlessly, users can choose their search method according to a condition such as the number of images to be searched, users' domain knowledge and characteristics of metadata assigned to images.
  • The new interface allows users to retrieve previously saved image collections by the name assigned by the users.
 

Students

  • Images selected by faculty are available on the web. Students can review images for assignments, exams and other tasks via the Web.
  • The personalization feature offers a flexible environment for students. They can change the contents and appearance of the image site to suit their particular needs (e.g. printing the page to prepare for an exam, reviewing images on the screen.)
 

Webmasters

  • HTML codes are dynamically generated based on the images faculty have selected by themselves. It eliminates time-consuming process of HTML generation and modification.
  ^ top
  4. Plan (Schedule)
 

User Studies and Task Analysis

We would like to design the system based on extensive user studies through interviews, observations and possibly surveys with three groups people on campus, faculty members, students and webmasters who are all involved with courses in which images play an important role. We are interested in understanding

  • how they use images (regardless of analog or digital) for educational purposes
  • problems they encounter when using images in courses
  • what they want to do when images are available on the web

Based on the user studies, we analyze the tasks that potential users (faculty and students) perform using digital images.
 

Prototyping

We plan to develop prototypes of user interfaces for

  • image selection/grouping
  • image access/site personalization

We start from paper-based low-fi prototype and then create computer-based interactive prototype for each interfaces.
 

Evaluation

At each stage of prototyping, we plan to conduct a series of evaluations of the interfaces. The design of the interfaces is expected to evolve through several iteration of design and evaluation. We evaluate our design by usability tests with potential users and other usability evaluation methods such as heuristic evaluation and surveys. We compare our design with other tools available for image selection and presentation such as Insight and SPIRO, based on user tests with potential users, interviews and classroom observations.

We also study image selection/presentation tools such as Insight and SPIRO currently available on campus
  ^ top
  5. Participating Students
 

Chaitee Sengupta | cks@sims.berkeley.edu

Masako Sho | shom@sims.berkeley.edu
  ^ top
last updated: 02/10/2000           i
shom@sims.berkeley.edu | cks@sims.berkeley.edu s e l e c t